The invention further relates to a fluorescence microscope for the three-dimensional imaging of a sample with a location accuracy exceeding the optical resolution, having: an illumination device which is designed for the purpose of repeatedly exciting fluorescence emitters in the sample to emit fluorescence, an imaging device having an imaging beam path with the optical resolution, designed for the purpose of producing still images of the sample at the optical resolution, a control device which is designed for the purpose of controlling the illumination device and the imaging device in such a manner that multiple still images of the sample are produced, wherein the fluorescence emitters are excited to emit fluorescence in such a manner that at least a subset of the fluorescence emitters in each still image is isolated in such a manner that the images of these fluorescence emitters can be separated in the still images within the optical resolution. The control device is also designed for the purpose of localizing the positions of the isolated fluorescing fluorescence emitters in the generated still images with a location accuracy exceeding the optical resolution, and generating a high-resolution composite image therefrom. The imaging beam path comprises a dividing element which divides each still image into a first and a second sectional image, wherein the first sectional image images a first focal plane in the sample, and the second sectional image images a second focal plane in the sample, and the imaging device includes a camera on which the two sectional images are imaged next to each other.
Various different methods have been developed in the prior art to overcome the diffraction limit in microscopy. A method, abbreviated as PALM (photo-activated light microscopy), is known from WO 2006/0127692 and DE 102006021317 A1, which uses a marking substance to image a sample, wherein said marking substance can be activated by means of optical radiation. The marking substance can only emit specific fluorescent radiation in the activated state. Inactivated molecules of the marking substance do not emit fluorescence radiation—or at least no noticeable fluorescence radiation, even after radiation with excitation light. For this reason, the excitation light is generally termed the switching signal. In the PALM method, the switching signal is applied in such a manner that at least some of the activated marking molecules are spaced apart from neighboring, activated marking molecules in such a manner that they are separated when viewed on the scale of the optical resolution of the microscope, or can be subsequently separated by image processing methods. In this case, one says that a subset of the fluorescence emitters have been isolated. After the fluorescence has been captured, the center of the radiation distribution for these isolated emitters is determined, said distribution being the result of the limit of the resolution. From this, it is possible to calculate the position of the molecules with higher precision than the optical resolution actually allows. This process is termed localization. The enhanced resolution resulting from a computational determination of the nucleus of the diffraction distribution is also termed “super resolution” in the technical literature in English. This resolution requires that at least a subset of the activated marking molecules in the sample can be differentiated—that is, isolated—at the optical resolution. Then, their position can be determined with a higher precision, and they can be localized.
To isolate individual fluorescence markers, the PALM principle exploits statistical effects. For a fluorescence marker which can be stimulated to emit fluorescence after receiving the switching signal at a given intensity, it is possible to adjust the intensity of the switching signal so that the probability of activating fluorescence markers present in a given area of the sample is so small that there is a sufficient number of sub-regions in which only fluorescence markers which can be differentiated within the optical resolution emit fluorescence.
The PALM principle has been further advanced with regards to the activation of the molecule which is targeted for detection. By way of example, for molecules which have a long-lived non-fluorescing state and a short-lived fluorescing state, a separate activation using activation light which is different in spectrum from the excitation light is not at all necessary. Rather, the sample is first illuminated with high-intensity excitation light in such a manner that the overwhelming majority of the molecules are brought into the long-lived state where fluorescence is not possible (e.g. a triplet state). The remaining molecules which are still fluorescing are thereby isolated with respect to the optical resolution.
It is also noted that the PALM principle has also been denoted in the technical literature with other abbreviations, such as STORM, for example. In this description, the abbreviation PALM is used for all microscope-based imaging which achieves a localizing resolution beyond the optical resolution of the apparatus being used, by first isolating fluorescent molecules and then localizing the same. The PALM method has the advantage that it is not necessary to have high localizing resolution for the illumination. A simple wide-field illumination is possible.
The PALM principle requires that many still images of the sample are captured, each containing subsets of isolated molecules. In order to image the sample as a whole, the number of the individual images in total must be sufficient to ensure that as many molecules as possible are at least present one time in one subset. The PALM method therefore regularly requires a plurality of still images, which requires a certain period of time for a composite image to be captured. A significantly complex calculation process is involved because a plurality of molecules must be localized in each still image. Large amounts of data are involved.
This location accuracy is only achieved laterally, by the localization in still images—that is, in a plane to which the image plane of the camera is functionally assigned. The methods are therefore limited in this respect to a two-dimensional analysis of a sample. The PALM principle is therefore combined with a TIRF excitation, which ensures that only fluorophores in a thin layer of the sample emit fluorescence.
Approaches are also known in the prior art for the localization of luminescing fluorescence markers in the third spatial dimension, which is the depth dimension with respect to the imaging of the sample. The term “depth dimension” in this case means the direction along the incident light path—that is, along the optical axis.
The publication B. Huang et al., Science 319, page 810, 2008 describes an imaging beam path for the PALM principle, wherein a weak cylindrical lens lies in the imaging beam path, thereby leading to a specific astigmatic distortion. As a result, the image of the marker on the camera is elliptically distorted as soon as the marker is positioned above or below the focal plane—that is, the symmetry point of the point spread function. The information on the depth position of the fluorescing fluorescence marker can be obtained from the orientation and the degree of the distortion. A disadvantage of this method is that the local environment and orientation of a molecular dipole can also lead to distortion of the image of the fluorescing fluorescence marker, and this distortion nevertheless has nothing to do with the depth position. Such fluorescing fluorescence markers therefore are assigned a false depth value, depending on their orientation.
The publication Pavani et al., PNAS 106, page 2995, 2009, suggests modifying the point spread function in the imaging process to give a double helix structure, by means of a spatial phase modulator. The one-dimensional images of individual, luminescing fluorescence markers then become double spots. Their depth position is encoded in the angular orientation of the common axis of the double spots.
According to the publication by Shtengel, et al, PNAS 106, page 3125, 2009, photons which are emitted by the fluorescing fluorescence markers are caused to interfere with themselves. For this purpose, two lenses which are assembled in the 4π configuration are used to simultaneously observe the fluorescing fluorescence markers. By means of a special, 3-way beam splitter, the different optical paths obtained in this manner are made to achieve interference. Each of the resulting three-point images is detected by a camera, and the proportional intensities of the three-point images provide information on the depth positions.
The publications Toprak et al., Nanolet. 7, pages 3285-3290, 2007, and Juette et al., Nature Methods 5, page 527, 2008, describe an approach wherein a 50/50 beam splitter is installed in the imaging beam path and splits the image of the sample into two sectional images which can be detected independently. In addition, an optical path length difference is inserted into one of the beam paths obtained in this manner, in such a manner that two object planes are produced from the two beam paths, which are spaced apart from each other in the z-—that is, depth—dimension by approximately half of the minimum optical resolution (for example 700 nm), or by the whole minimum optical resolution. The depth position of fluorescence markers which lie between these two planes is then obtained by analysis of the two sectional images of the same fluorescence marker (for example with respect to the width of the point spread function, which can be analyzed by a type of subtraction), or by a corresponding fitting of a three-dimensional point spread function. The method requires two highly resolved sectional images and a precise adjustment of the beam paths and calibration measurements in order to achieve a superimposition of these two sectional images with sub-pixel precision. In addition, the two sectional images of a fluorescence marker generally have a different shape because the lateral expansion of the point spread function of a system being imaged changes according to the position of the object plane being observed.
DE 102009060490 A1, which also lists further citations of literature on 3D high-resolution, likewise follows the approach according to Toprak et al. in this class, to split the image of the sample into two images.